Explaining Group Influence: The Role of Identity and Emotion in Political Conformity

and Polarization
Author(s): Elizabeth Suhay
Source: Political Behavior , March 2015, Vol. 37, No. 1 (March 2015), pp. 221-251
Published by: Springer

Stable URL: https://www.jstor.org/stable/43653424

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Springer is collaborating with JSTOR to digitize, preserve and extend access to Political
Behavior

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms
Polit Behav (2015) 37:221-251
DOI 10.1 007/s 1 1 1 09-0 1 4-9269- 1

ORIGINAL PAPER

Explaining Group Influence: The Role of Identity

and Emotion in Political Conformity and Polarization

Elizabeth Suhay

Published online: 8 March 2014

© Springer Science+Business Media New York 2014

Abstract Evidence has accumulated that people often conform to political norms.
However, we know little about the mechanisms underlying political conformity.
Whose norms are people likely to follow, and why? This article discusses two
phenomena - social identity and "self-conscious" emotions - that are key to
understanding when and why people follow the crowd. It argues that adherence to
in-group norms is a critical basis of status among in-group peers. Conformity
generates peer approval and leads to personal pride. Deviance generates disapproval
and causes embarrassment or shame. These emotional reactions color an individ-
ual's political perspectives, typically generating conformity. These same mecha-
nisms can spur between-group polarization. In this case, differentiation from the
norms of disliked out-groups results in peer approval and pride, and conformity to
out-group norms disapproval and embarrassment or shame. This framework is
supported by the results of two experiments that examine the influence of group
opinion norms over economic and social aspects of citizens' political ideologies.
One exogenously varies the social identity of attitudinal majorities; the other primes
the relevant emotions. In addition to contributing to the study of political conformity
and polarization, this article adds to our growing understanding of the relevance of
social identity and emotion to political life.

Keywords Conformity • Polarization • Social identity • Emotion • Norms •

Political opinion

As of July 1, 2014, the author will be Assistant Professor of Government at American University.

E. Suhay (El)
Department of Government and Law, Lafayette College, 100 Kirby Hall, Easton, PA 18042, USA
e-mail: suhaye@lafayette.edu

Ô Springer

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms
 222 Polit Behav (2015) 37:221-251

After a long hiatus, political scien

influence. Researchers in this are
decisions are not made in a social v
the words and actions of others. M
focus on one of two topics: pol
Huckfeldt and Sprague 1995; M
2011) or conformity to social no
1998; Nadeau et al. 1993; Nickers
accumulated findings from stu
influence one another's political a
While scholars have successfully
settings and with respect to man
attention has been paid to the me
political decisions are affected by
rational choice theories (e.g., Do
that influence between discussant
argument (e.g., Huckfeldt and Spr
be appropriate for the study of po
social norms. In fact, the conform
defined by the absence (or irrel
argument.
What fills this mechanistic void? Research in psychology and sociology suggests
that identification with peers (Turner et al. 1987, 1991) and "self-conscious"
emotions - especially pride, embarrassment, and shame (Lewis 2000; Scheff
1988) - are worth consideration. While some studies of social influence in political
science have discussed these mechanisms and tested them indirectly (e.g., Gerber
et al. 2010; Sinclair 2012), surprisingly few studies have provided persuasive
evidence of their causal role in influence by directly manipulating them
experimentally and/or measuring them via participant self-report. The result is that
speculation as to the role of identity and emotion in politically relevant social
influence far outweighs the evidence. In addition, while some political scientists
have discussed these mechanisms independently, none have combined them into an
integrated theoretical framework.
The framework presented here focuses on the mechanisms underlying conformity
to social norms1 and argues that self-conscious emotions encourage individuals to
adopt the norms of groups with which they closely identify.2 More specifically,
pride stemming from perceived approval of in-group peers signals to the individual
that a conforming viewpoint or action is valid, whereas embarrassment or shame
stemming from perceived disapproval of in-group peers generates doubt. The result

1 While political scientists often associate norms with behavior, group norms may be behaviors, thoughts,
or even feelings that are typical within a group (Hogg and Reid 2006, p. 8).
Many types of groups are relevant, ranging from face-to-face (e.g., workplace, neighborhood) to
broader demographic (e.g., race, sex, religious) groups. A person who identifies with a group perceives it
to be an important element of his or her personal self-concept (Tajfel and Turner 1986).

<0 Springer

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms
Polit Behav (2015) 37:221-251 223

is conformity to in-group norms

group polarization as well. Individ
norms in part because they perce
actions that are different from th
that resemble disliked out-groups
that examine the role of identity
norms over economic and social
the evidence supports the theoreti
This article contributes to rese
political conformity and polarizatio
the political relevance of emotion
accumulated, the role of emotions
as has the political relevance of th
pride, embarrassment, and sham
mysterious phenomenon of share
science that political opinions and
social-psychological dynamics desc
why this is the case.

Theoretical Framework

Social Influence Over Political Opinion and Behavior

Over a half century ago, the Columbia and Michigan schools of voting behav
were quite taken with social influence and, in particular, the idea of socia
conformity. Berelson et al. (1954, p. 122) argued that "[d]uring a campaign politi
preferences are 'contagious' over the range of personal contacts." Campbell et al.
(1960) showed that voting preferences of citizens mirrored those of their racial,
religious, and union affiliations, even after controlling for characteristics th
citizens were likely to share with fellow group members (e.g., income, education
occupation). The authors attributed these findings to forces less "rational" th
ordinary persuasion. Berelson et al. argued that "[f]or many voters politic
preferences may better be considered analogous to cultural tastes," with origins
ethnic, sectional, class, and family traditions (1954, p. 311). Campbell et al. wrot
"[N]orms and values attributed to a generalized 'group': these are the expectation
concerning appropriate behavior for [e.g.] the 'loyal' Catholic or union member.
is the group standards that are psychologically real and are responsible for influen
when it occurs" (1960, p. 296).
These voting scholars had been influenced by the findings of experiment
psychologists studying group dynamics who, in the preceding decades, had been
working to understand a particular social influence phenomenon, variously label
"group pressure," "conformity," or "majority influence." For example, She
([1936] 1966) demonstrated with his "auto-kinetic effect" studies that a group of

£) Springer

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms
 224 Polit Behav (2015) 37:221-251

previously unacquainted individual

influencethe individuals' judgment
Asch (1951), in his line-length s
experimenter, via confederates,
judgments even when they we
experimental studies followed, m
and political norms (see Turner 19
that individuals frequently change
majority3 even when no accompany
given.
However, the reasons for conformity remained murky, even among psycholo-
gists. The lack of a clear mechanism of influence cast doubt on the conformity
phenomenon. Coupled with an overall decline in interest in social influence,
political scientists seemed to quickly lose interest in the topic of conformity. When
the study of politically relevant social influence reemerged several decades later,
interest had shifted to discussion networks (e.g., Beck 2002; Huckfeldt and Sprague
1987, 1995; McClurg 2006). One benefit of this influence paradigm is that it fits
more comfortably within standard rationality assumptions. Scholars in this tradition
tend to argue that citizens exchange - and are persuaded by - information and
logical arguments.
Until very recently, only a handful of contemporary studies had tackled the
subject of political conformity. The most prominent are Mutz' s Impersonal
Influence (1998) and Noelle-Neuman's Spiral of Silence (1993). Although each
builds on the conformity research tradition in psychology, these works differ from
that tradition in important ways. For example, similar to social network scholars,
Mutz fits conformity into a rationalistic framework; she argues that some people
construct arguments as to why the majority believes as it does, persuading
themselves in the process. Mutz' s framework represents an alternative, although
ultimately complementary, explanation for political conformity to the one offered
here. Noelle-Neuman moves in a different direction. She explains her "spiral of
silence" findings by arguing that people are motivated to avoid embarrassment;
however, she never directly tests this assumption.
Interest in conformity has noticeably increased more recently in political science,
with scholars quickly amassing considerable evidence that peers often influence one
another's political opinions and behaviors even when substantive information
exchange and arguments are absent (e.g., Bolsen 2013; Gerber et al. 2008; Gerber
and Rogers 2009). While this more recent scholarship has focused more on
demonstrating the influence of social norms than on investigating the mechanisms

3 This article focuses on descriptive norms (an opinion held or behavior engaged in by the majority), not
injunctive norms (opinions or behaviors considered socially desirable or even morally correct). Note,
however, that the line between descriptive and injunctive norms is unclear. Like injunctive norms,
descriptive norms typically suggest to people how they ought to behave (Theiss-Morse 2009), signaling
opinions and behaviors "appropriate" for group members (Turner et al. 1987). Also similar to injunctive
norms, descriptive norms can be enforced via social-psychological rewards and sanctions (Scheff 1988).

Springer

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms
Polit Behav (2015) 37:221-251 225

that underlie it, scholars have beg

with those proposed in this article
Sinclair (2012) argues that socia
influence (also see Alwin et al.
demonstrates that neighborhood c
vote than outsiders. One limitation
manipulated only roughly (assum
to Noelle-Neuman (1993), 4 Ger
argue that the desire to feel prid
citizens to conform to voting nor
people are more likely to vote wh
evidence that this type of experim
persuasive, the emotions assum
manipulated or measured, leaving
sum, while strong evidence for p
have named social identity and pa
research demonstrating a causa
manipulation and/or careful meas

Explaining Conformity and Polari

The primary goal of this article is

political conformity, with a secon
polarization.5 Psychologists and
these phenomena, incorporating th
prominent theoretical framework
polarization.

The Social Identity Perspective

This article draws important insig

categorization theory" (SCT) (Tu
upon - and expands the scope of -
in conjunction with Turner (e.g.,
political scientists are more famili
on SCT for two reasons. First, un

4 See also work by Elster on adherence t

5 Many definitions of polarization exist. In
the norms of two or more groups increasi
6 The borders of social identity theory are
literature. A common error is to attribute
self-categorization theory to social ident
Haslam et al. 2010). Such confusion is und
authorship (Tajfel and Turner). And, in s
label to refer to both theories, while clarif
Huddy 2001; Huddy and Khatib 2007). In th
umbrella term to refer to both theories to

â Springer

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms
 226 Polit Behav (2015) 37:221-251

specified theory of the psycholo

intrag roup, as well as intergrou
been used to explain a wider ra
identity theory. Most notably, SC
for social conformity and polari
et al. 2010).
Self-categorization theory arg
between-group polarization is s
group membership into his or h
When a relevant social identity i
group members (Turner 1985; Tu
characterized by a high "meta-co
not only resemble in-group mem
out-group members. Numerous e
1991). People tend to conform to
not out-groups (Abrams et al.
Further, taking into account the
group members who identify mor
more influenced by its norms on
also see Theiss-Morse 2009). Final
group, particularly a derogated one
opposite direction (Clark and M
These researchers, by employin
isolated social identification as on
as a cause of polarization). Thus,
causal influences on within-group
affiliate with like, see Lazer 201
What is it about identificatio
Psychologists have long argued
"social reality tests" when peers
point of view (Festinger 1950; Tu
identifies provide the best social
agreement of similar others in t
function of the objective world
members generates "subjective v
them generates "subjective inval
that the logic is reversed for out
generate subjective invalidity.

7 These studies do not argue that homoph

argue that group influence over the indivi
but also that their effects can be disentan

£) Springer

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms
 Polit Behav (2015) 37:221-251 227

Self-categorization theory offers

mity and polarization; however, it
emotion.8 Empirical accounts of p
with emotion, especially when peo
Milgram 1992; Noelle-Neuman 1
formulated by Turner - says little
considers emotion to be highly rel
In particular, a key tenant of so
achieve or maintain positive self-es
from a comparison outgroup on s
see, e.g., Tajfel and Turner 1986).9
self-esteem - relevant to social influence?

Self-Conscious Emotions

A group of emotions called "self-conscious emotions" - especially embarrassment,

shame, and pride - deserve our attention in seeking to understand political
conformity and polarization. Not only are they closely linked to individuals' levels
of self-esteem, but they also have been explicitly named by many researchers as
responsible for social conformity.
Although he does not explicitly discuss emotion, Turner argues that those who
achieve a high meta-contrast ratio (again, not only resembling the in-group, but also
being distinct from the out-group) are evaluated most highly by in-group peers
(Turner 1985; Turner et al. 1987; also see Theiss-Morse 2009). According to
other researchers, those who exemplify shared community ideals tend to be treated
with respect and admired (Lewis 2000), whereas norm breakers are often subject to
derogation (or shaming) (Marques et al. 2001). While the former case is likely to
lead to positive self-esteem and pride, the latter tends to cause embarrassment and
shame10 (Asch 1951; Lazarus 1991; Lewis 2000; Milgram 1992). The result is often
conformity at the individual-level to social norms and values (Lewis 2000; Scheff
1988). This phenomenon has been described most fully by sociologist Thomas
Scheff: "[P]ride and shame make up a subtle and pervasive system of social
sanctions. This system leads to experiencing social influence as compelling" (Scheff
1988, p. 396). The system functions continuously, even when we are alone, because

Defining emotion is notoriously difficult. Lazarus (1991) says that "emotion is an integrative,
organismic concept that... unites motivation, cognition, and adaptation in a complex configuration" (40).
Cognitive appraisals of whether and how a situation is relevant to an individual's goals set in motion
(ideally) adaptive action tendencies and coping mechanisms. Much of this psychological and
physiological activity occurs automatically and subconsciously, but some may be conscious, including
subjectively felt "feelings."
9 Self-categorization theory also includes this idea, but Turner casts it in cognitive terms (Turner 1985,
p. 261) and does not incorporate it into his explanation for conformity and polarization. Turner's de-
emphasis of the self-esteem plank of social identity theory may stem from uneven empirical support for
this proposition (see Brown 2000). The self-esteem hypothesis discussed in this article is related but
clearly distinct from that discussed as a part of social identity theory.

10 While Lazarus argues that the main difference between these two emotions is intensity level (Lazarus
1991), note that some argue for differences in kind as well as degree (e.g., see Miller 2007).

Ô Springer

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms
 228 Polit Behav (2015) 37:221-251

we can imagine others' reaction

Scheff 1988).
The explanation for conformity
compelling but also underspecifie
every majority norm of whic
Americans are unlikely to feel em
even if made aware that many
Democrat is unlikely to gloat wh
that her opposition to gay marr
While the literature on self-cons
conformity occurs in respons
(suggesting social identity is r
generally untested empirically.

An Integrated Perspective on Soc

It seems clear that an integrat

emotions would benefit the st
influence more specifically. Whi
direction - intergroup emotion
esteem (Luhtanen and Crocker 19
dynamics (e.g., prejudice) than th
conformity.11 This should not b
the extent to which group bound
been understudied.

Thus, this article draws directly from the self-categorization and self-conscious
emotions literatures. Focusing on their applicability to political variables, a model
of conformity and polarization is introduced. In group contexts, interpersonal
comparative judgments are common among in-group members (Turner 1985,
p. 255). The basis of much of this comparison is the extent to which individuals
conform to in-group norms and differ from out-group norms, particularly norms
held by derogated out-groups. In other words, the extent to which individuals
resemble in-group prototypes is a key basis of status within a group. Prototypicality
(in general or in a specific domain) tends to generate pride in individuals because
they perceive respect and approval from in-group members; non-prototypicality

1 1 Intergroup emotions theory, a broadly applicable theory developed by Smith and Mackie (see, e.g.,
Mackie et al. 2009), argues that, when social identity is salient, people will appraise situations and
experience relevant emotions in accord not with their personal identities but, rather, with their social
identities. For example, when a social identity is highly salient, an out-group attack on the in-group is
experienced as an attack on the self, and fear or anger directed at the out-group is generated as a result.
Luhtanen and Crocker (1992) provide an updated take on social identity theory's "need for positive
distinctiveness" with their construct "collective self-esteem," i.e., that part of an individual's self-esteem
that is derived from the status of one's in-group(s) within society at large. This concept is distinct from
what they call "membership esteem," that part of self-esteem stemming from one's status as an individual
within the group. While collective self-esteem may be the cause of many important intergroup
phenomena (including prejudice and people's desire to "exit" low status groups), emotions scholars have
made clear that within-group conformity hinges on membership esteem.

£) Springer

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms
 Polit Behav (2015) 37:221-251 229

tends to generate embarrassment

respect and disapproval. Note t
communicated explicitly or ma
individual.

These self-conscious emotional reactions color an individual's perspective. When

people perceive in-group approval in response to their adherence to in-group norms
(or their differentiation from out-group norms), the resulting pride strengthens their
internal commitment to in-group norms. When people perceive in-group
¿//¿approval in response to their deviance from in-group norms (or adherence to
out-group norms), the resulting embarrassment or shame weakens their commitment
to those deviant attitudes or behaviors. In this way, norms shared by a majority of
in-group peers often become internalized over time by group identifiers.12 Finally,
note that these conforming and polarizing influences are limited to the in-group.
Evaluative feedback from individuals outside the boundaries of the in-group is
largely orthogonal to a person's self- worth and, thus, his or her self-conscious
emotional experiences.
Given the greater relevance of in-group status to self-worth among high
identifiers, these mechanisms should be more pronounced among high identifiers,
leading to greater conformity and polarization among them. Although for the most
part outside the empirical scope of this article, note that the likelihood and extremity
of conformity and polarization will also vary according to additional factors,
including personality traits, group cultures, and social contexts that may increase or
decrease the salience of particular group prototypes, the tendency to make
interpersonal comparative judgments, or the extremity of emotional reactions to
perceived judgments.
In sum, the theory - which I refer to as "social-emotional influence theory" -
provides one explanation for why political conformity occurs. It also argues that
these same mechanisms that underlie conformity to in-group norms underlie
between-group polarization. In other words, it is argued that polarization - while
admittedly dependent on the existence of two groups - is at least in part motivated
at the individual level by intrag roup social-psychological processes. While the
theory is broadly relevant to political conformity and polarization, this article
narrows the subject of empirical inquiry in two respects. First, it investigates
"impersonal" conformity, as opposed to conformity to individuals with whom one
is interacting in person. Impersonal conformity represents a more difficult test of the
theoretical framework, as in-group peers are distant from study participants. Second,
it focuses on conformity to opinion, not behavioral, norms. This allows for the
elimination of certain confounding factors, such as a calculated concern for one's
reputation due to deviance that is easily observed by peers. In addition, a focus on
group influence over private opinion rather than public behavior (or opinion

12 Many distinguish between "informational influence" ("true" influence based on information) and
"normative influence" (surface compliance due to an effort to ingratiate oneself with peers) (Deutsch and
Gerard [1955] 1965) but others have challenged this dichotomy (e.g., Turner et al. 1987; Hogg and Reid
2006). This article suggests that what many would call "normative influence" can in fact cause the "true"
influence often associated with informational influence.

â Springer

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms
 230 Polit Behav (2015) 37:221-251

expressed in public) may carry m

has been internalized.

Hypotheses

This theory results in six key hypotheses regarding political influence that are tested
below:

In- group conformity (HI) When encountering a new or unexpected in-group

norm, individuals will shift their political views in the direction of conformity with
the majority.
Out- group polarization (H2) When encountering an out-group norm, individuals
will not change their views in the direction of conformity with the (out-group)
majority; and, if the out-group is disliked by their in-group peers, individuals will
shift their views away from those of out-group members.
Identity strength and conformity (H3) Due to the accumulation of identity-based
influence over time, in-group identification will be positively correlated with
adherence to in-group opinion norms.
Identity strength moderation (H4) The influence patterns described in HI and H2
will be moderated by the strength of individuals' in-group identification.
Emotional reactions to group norms (H5) Agreement with the in-group majority
generates pride among individuals, and disagreement embarrassment/shame;
agreement with a derogated out-group majority generates embarrassment/shame
among individuals, and disagreement pride.
Emotion moderation (H6) When encountering group norms, individuals who
experience elevated levels of pride or embarrassment/shame are more likely to
change their political views (in the direction of the in-group or away from the out-
group, as relevant).13

Experimental Evidence

Given that experiments are the gold standard for assessing causation (Morton and
Williams 2010), two experiments were designed to test these mainly causal
hypotheses. The first experiment focuses on religious group influence over citizens'
stands on social issues and emphasizes the role of social identity. Citizens' opinions
on issues such as abortion and gay marriage depend to a significant extent on their
religious faith (Leege et al. 2002). To the extent that religious affiliation represents a
social identity (Djupe and Gilbert 2008; Wald et al. 1988), individuals' opinions on
these issues are likely to be affected by the opinions held by religious groups - both

13 It is common for causal variables to be both mediators and moderators in a theoretical framework. This
is the case with respect to self-conscious emotions in the model proposed herewith (see the "An
Integrated Perspective on Social Influence" section). Because the critical test of emotions' causal
influence in Study 2 below involves exogenous arousal, a moderation test is performed. However, one
could argue that - taken together - the two studies test mediation: Group norms arouse self-conscious
emotions in Study 1, and their causal effects are demonstrated in Study 2 (see Bullock et al. 2010; Imai
et al. 2012).

Ô Springer

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms
Polit Behav (2015) 37:221-251 231

in-groups and disliked out-group

conformity with respect to socia
role of self-conscious emotions. In
evoke the value economic indi
American political culture (Feldm
college peers, a highly relevant s
(Alwin et al. 1991; Moskalenko
Taken together, these two expe
causal agent suggested by the t
exogenously varied in one experi
Second, group influence is examin
to economic and social policy, i.e.,
of the variance in political ideolog
are not formal political groups. As
likelihood that these groups influ
people choose to affiliate with th
Sprague 1995). This represents a n

Study 1 : Religious Identity and

Study Sample

The study was conducted with 2

large Archdiocese in the Midwe
bulletins and announcements from
a short, anonymous, on-line s
University of Michigan researche
$5 donation to a Catholic charity
Most participants had a high lev
the pre-test question, "How imp
and 44 % said "extremely." Most
conservative values, with 71 %
traditional church teachings on
other sample characteristics, s
(69 %), likely reflect the church-

Study Design

The experiment had a 1 x 4 betw

study by filling out a pre-test th
and demographics. Participant

14 While a religious person's opinions on

leaders, etc., the focus here will be specif
15 Reflecting the demographic attribute
homogeneous (98 % white) and upper-in
Catholics as a whole. (National statistics

â Springer

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms
 232 Polit Behav (2015) 37:221-251

treatments or the control grou

exogenously varied by exposing so
group and others to the views of
More specifically, those in the "C
a majority of American Catholics
marriage. Those in the "Catholics
contrary, most Catholics believ
conservative on Church teachings
in the "Evangelicals are conservat
socially conservative views on divo
received no such information. Inf
reputable public opinion surveys
Center 2008). The Catholics are con
presenting responses to different s
"Evangelicals are progressive" stim
that could be interpreted in this fa
Following the stimulus, partici
included questions on social and p
then were debriefed.

Measures

The dependent measure is a socially conservative attitude scale made up of nin

questions touching on five (equally weighted) salient themes - divorce, pre-marita
sex, homosexuality, sexuality and teens, and abortion (a = .83). Variation in
participants' strength of identification with other Catholics was measured in the pre
test with an additive scale made up of two questions adapted from Huddy an
Khatib (2007) (r = .71). 17 Emotional responses to the stimuli were assessed at the
end of the study with two questions that asked treated participants whether they fe
pride or shame in response to the opinion information (i.e., the stimulus). A
variables range from 0 to 1, with 1 representing the most conservative, identified, o
emotional response. See Appendix 1 for exact wording and response options.

Data Analyses

Expectations with respect to the main treatment effects are as follows. The general
conformity hypothesis (HI) is tested with the "Catholics are progressive"

16 American Evangelicals and Catholics tend to be mutually exclusive groups (Putnam and Campbell
2010). Only one participant, removed from the sample, identified as Evangelical. In the post-test,
participants rated Evangelicals 25 points lower than Catholics on average on a 100-point scale. (Ratings
were not influenced by experimental treatments.)
The questions that make up the identity measure reflect the social identity perspective's definition of
"identification": "the extent to which the category is valued and contributes to an enduring sense of self'
(Haslam et al. 2010, p. 349). By design, the measure is somewhat out of sync with self-categorization
theory's exclusively cognitive focus (the "importance" question has an affective component). However,
the measure admittedly does not emphasize affiliative, emotional attachments to the same extent as those
of some authors, such as Theiss-Morse (2009).

<0 Springer

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms
 Polit Behav (2015) 37:221-251 233

treatment; participants in this co

values in the post-test than the co
tested by comparing reactions t
icals are conservative" treatments
least as conservative as the control
more progressive than the cont
participants.
Figure 1 displays the means and 90 % confidence intervals19 for the four
experimental groups with respect to social conservatism. The patterns are as
predicted. Relative to the control group, participants who learned that most Catholics
were unsupportive of conservative Church stands were more progressive in the post-
test. And, while participants who learned that other Catholics were socially
conservative were equivalent to the control group, those who learned that Evangelicals
were socially conservative were more progressive than the control group.
We can assess significance levels using OLS regression. The Social Conserva-
tism Scale was regressed onto dummy variables representing the three treatment
groups; the control group was excluded as the comparison group. The model is as
follows: Social Conservatism = ß0 + ßiCatholics Conservative -|- ß2Catholics
Progressive + ß3Evangelicals Conservative + 8.20 The coefficients for the "Cath-
olics are progressive" ( b2 = -.077) and "Evangelicals are conservative" ( b3 =
-.070) conditions are both statistically significant (p < .05 one-tailed).21 See the
first column of Table 1.
We turn next to evaluating the hypotheses related to identity strength. With
respect to H3, we would expect there to be a positive relationship between Identity
Strength and Social Conservatism given the longstanding association between the
American Catholic community and opposition to abortion, gay marriage, and the
like. To avoid contamination by the experimental treatments, this hypothesis is
evaluated by analyzing the control group only. With respect to H4, we expect strong
identifiers to be more influenced by the "Catholics are progressive" and
"Evangelicals are conservative" stimuli than weak identifiers. To test these
hypotheses, Identity Strength is added to the above equation along with interaction
terms between Identity Strength and each of the three treatment groups. The results
are in the last column of Table 1.
The coefficient on Identity Strength , which represents the relationship between
this variable and the dependent variable in the excluded control group, is very large
and highly significant ( b = .611, p < .001). On average, the strongest Catholic
identifiers were 61 percentage points higher on the Social Conservatism Scale than

18 Because the information does not challenge de facto assumptions, it is unlikely to influence
participants.
19 Note that confidence intervals surrounding two estimates that are significantly different from one
another may still overlap somewhat. (Confidence intervals are wider than standard errors.)
20 Statistical tests assessing experimental group balance on demographic and political variables showed
that randomization was successful and, therefore, no control variables were used.

21 One-tailed tests are employed throughout this section given the directional nature of the hypotheses.
Note p-values for the following additional contrasts: Catholics conservative vs. Catholics progressive
(p = .07); Catholics conservative vs. Evangelicals conservative (p = .08).

Ô Springer

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms
 234 Polit Behav (2015) 37:221-251

* { i i ;!
IlilllllllllllWllllBlllllilWIIIIIIIllllllllllillllllllllllllllllllllllllllllllllllllWIllllllllllllllllllllllgiWilBlliillll

S s S
Fig. 1 Social Conservatism Scale mean
experimental group means and 90 % con

Table 1 Effects of treatments on partic

Social conservatism Social conservatism

(w/identity terms)

Constant .619*** (.032) .192** (.073)

Catholics conservative -.013 (.046) .035 (.123)
Catholics progressive -.077* (.043) -.094 (.107)
Evangelicals conservative -.070* (.043) .026 (.110)
Identity strength - .611*** (.098)
Catholics conservative x Identity - -.100 (.162)
Catholics progressive x Identity - .007 (.142)
Evangelicals conservative x Identity - -.148 (.148)
N 209 206

Table entries
*** p < .001

the lowes
hypothesi
zero, agai
tity Stren
misses sta

â Springer

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms
 Polit Behav (2015) 37:221-251 235

Fig. 2 "Evangelicals conservative" treatm

predicted values for experimental groups
level. Values calculated and graphed usin

is graphed along with its confide

variable. While the treatment eff
of identity, it is significant at th
for H4.22
Finally, we can turn to the emotion measures to test H5. We would expect typical
sample members - who began the study indicating they were socially conserva-
tive - to feel the most pride in the "Catholics are conservative" condition, as they
reflect on the fact that their values are in line with a majority of American Catholics,
and less pride in the other two conditions, perceiving their values to be out-of-step
with a progressive majority or in sync with a disliked out-group. The mirror
opposite pattern is expected for shame, with the least shame expected in the
"Catholics are conservative" condition and more shame in the other two conditions.
Note that, all else equal, people tend to express much more pride than shame;
therefore, contrasts among treatment groups are examined for each emotion
separately.
The expected patterns emerge in Fig. 3. Difference-in-means tests were
conducted and most of the relevant comparisons reach statistical significance. With
respect to pride, the difference is greatest between those in the Catholics vs.

22 Similar results are obtained if this treatment effect is estimated separately for those with Catholic
identities above vs. at/below the scale midpoint. Those with strong Catholic identities appeared to shift
their views in the progressive direction in response to the "Evangelicals are conservative" stimulus
(b = - .126, p < .01) but weak identifiers did not.

Ô Springer

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms
 236 Polit Behav (2015) 37:221-251

Fig. 3 Average levels of pride and sham

Evangelicals conservative cond

emerges for the Catholics are co
respect to shame, those in the
experienced none whatsoever;
Catholics are progressive (p <
conditions.23

Study 1 Discussion

Overall, these findings support the relevant hypotheses. Catholics who learned that
other Catholics held progressive family values expressed more progressive views in
the post-test than the control group, and Catholics who learned that Evangelicals
were socially conservative also expressed more progressive views in the post-test.
These findings are especially compelling when one considers that most practicing
Catholics are respectful of a conservative Church hierarchy and likely less open to
progressive change than others, and that the conformity and polarization described
occurred in a sample with a median age of 54, an age when values and attitudes are
at their most stable (Sears and Levy 2003). Also, as expected, the strongest Catholic
identifiers in the sample were overwhelmingly more likely to be socially
conservative. It is all the more intriguing, then, that high identifiers appeared to

23 The total N available is too small for a test of H6 (emotion moderation). The reason for the small N is
as follows: (1) For methodological reasons, the analysis cannot include the control group (because the
emotion questions asked for reactions to the treatments) or participants who said in the pre-test that they
disagreed with socially conservative Church teachings (a different emotional pattern is expected from
such individuals, and there are too few to analyze separately). (2) There was significant attrition prior to
the emotion questions because they followed a difficult screening question at the end of the study. This
attrition is statistically unrelated to treatment group and, thus, does not threaten the causal inferences.
The final A is 31.

Ô Springer

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms
 Polit Behav (2015) 37:221-251 237

be more likely than low identifiers

exposed to conservative out-group
In addition, the data support
automatically evoked when group
study participants who expressed s
study felt the most proud and t
Catholics supported those values (as
more ashamed when they learned
most Evangelicals supported them
causal role for these emotions in t
turn to a second experiment th
embarrassment.

Study 2: Self-Conscious Emotions and Economic Individualism Experiment

Study Sample and Design

Approximately 100 participants were recruited from the University of Michigan

Introductory Psychology subject pool.25 Students participated in person, in groups.
Each student received a paper-and-pencil questionnaire containing a pre-test,
stimulus, and post-test. Students were debriefed regarding the study aims and
methodology after completing the study.
The experiment had a 2 x 2 between-subjects design with random assignment of
treatments. Two factors were varied: information regarding whether most fellow
University of Michigan undergraduates supported or opposed economic individu-
alism, and whether or not participants' self-conscious emotions were primed. The
stimuli included the following. With respect to the opinion information, half of
participants read poll results that most University of Michigan undergraduates
supported economic individualism; the other half read poll results that most
University of Michigan undergraduates opposed economic individualism. All poll
results were fabricated for the purposes of experimental control.26 With respect to
emotion priming, approximately half of the participants read emotional scenarios
intended to prime either pride or embarrassment. (Embarrassment was chosen over
shame for ethical reasons.) In accord with standard practice in psychology, the
emotion primes preceded the opinion stimulus (e.g., see Small and Lerner 2005;

24 The identity moderation hypothesis (H4) was only partially supported, however; strong identifiers
were not more likely than low identifiers to shift in a progressive direction when exposed to progressive
m-group norms. One explanation for this null result is that some of the most devoted Catholics, who also
tended to be the most conservative, may have dismissed the progressive Catholics depicted in the study as
not "true Catholics." Borrowing again from Turner (1991), if an in-group norm is too different from a
person's personal beliefs, he or she may choose to redraw group boundaries - separating him or herself
from the former in-group - rather than conform. Exploring when such identity redefinition occurs is an
important topic for further study.

25 Demographics are as follows: 47 % Democratic, 23 % Republican, and 31 % Independent or "other."

78 % white, 8 % African American, 7 % Asian, 5 % Hispanic, and 2 % Native American. Men made up
52 % of the sample. The mean age was 19.
26 A post-test probe did not turn up any skepticism with regard to the veracity of these stimuli.

Ô Springer

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms
 238 Polit Behav (2015) 37:221-251

Table 2 Experimental design of economic

No emotion Emotion primed

(pride or embarrassment)

Majority supports economic individualism 23 20

Majority opposes economic individualism 22 20

Zajonc 1980). Primes were presented in the form of a quest

awareness of the manipulation and were unrelated to economic in
well as to politics in general) to decrease the likelihood of confou
See Table 2 for the basic design. The stimuli are available in Appe
The logic of how the opinion and emotion stimuli were expected
requires some explanation. First, it was expected that nearly all p
support economic individualism in the abstract (as it is a very po
US). Thus, the experimental design assumes participant commitme
value. Two Likert-type statements at the start of the questionna
should strive to be financially self-reliant" and "one ought to wor
same statements used in the opinion stimulus) - effectively scree
percentage of participants (15 %) who did not support economic
Thus, all analyzed participants who received the pro-individu
learned they were in the majority; those who received the anti-i
results learned they were in the minority. Second, the emotion pr
to these specific circumstances: those in the pro-individualis
assigned an emotion prime read pride-evoking scenarios, intende
likelihood that they would feel pride after finding themselves in
those in the anti-individualism condition also assigned an em
embarrassment-evoking scenarios, intended to increase the lik
would feel embarrassment after finding that their opinions were

Measures

The dependent measure is the average of responses to two questions on governmen

social welfare policy which heavily implicate the value economic individualism
The first question asked participants whether the federal government ought to giv
college scholarships to all low-income high school graduates or only to those who
have taken rigorous courses in high school. The second asked participants whethe
the federal government should provide welfare benefits to all low-income single
mothers or only to those who are willing to work for those benefits. (See Appendix

In other words, "incidental affect" (orthogonal to study content) rather than "integral affect" (arisi
in response to related content) is examined (see Blanchette and Richards 2010). The former allows one t
more cleanly isolate the causal influence of emotion on the dependent variable; the latter is usua
intertwined with cognitive content related to the study.
7.8 X .. , -, , . • . r . ii Ti .1
in most cases, tnese individuais expressed mild opposition to just one or tne statements, wnne mere
are too few cases for separate analysis here, note that adding these individuals to the analyses that follow
does not considerably alter the results.

Ô Springer

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms
 Polit Behav (2015) 37:221-251 239

for exact question wording.) Answ

additive scale Social Welfare Polic
that relatively individualistic answ
format of both questions was mod
National Election Studies.

The post-test questionnaire also included measures of participants' feelings. A

battery of emotion items was asked following the opinion stimulus to provide a
secondary test of emotional moderation. The instructions read: "Describe how you
feel right now by indicating the extent to which you feel each emotion." Answer
choices, arrayed on a five-point scale, ranged from "not at all" to "extremely."30
The emotions were listed in alphabetical order. These questions were used to create
a Self-Conscious Emotion Intensity scale, also placed on the 0 to 1 interval, as well
as another emotion measure described in the Discussion and Conclusion. The Self-
Conscious Emotion Intensity scale is scored to reflect responses in the "pride"
family ("proud" and "self-confident") for participants in the pro-individualism/
majority treatment groups and responses in the "embarrassment" family ("embar-
rassed" and "insecure") for participants in the anti-individualism/minority treat-
ment groups. This variable construction reflects the previous logic of the emotion
primes, where the specific emotion prime depended on whether the participant is in-
step with the majority (pride primed) or out-of-step with the majority (embarrass-
ment primed).31

Data Analyses

2x2 experiments are most easily analyzed and interpreted with ANOVAs, which
allow for straight-forward interpretation of the two "main effects" as well as the
interaction effect. Two separate ANOVAs were conducted - one with the emotion
treatment variable and one with the measured emotion variable.32 The experimental

29 Patterns of results presented below are similar when the variables are assessed separately.
30 This battery of questions was based on a widely-used emotion measure called the "Profile of Mood
States" (POMS) created by McNair and Droppleman (1971).
31 Note that it is inappropriate to test the moderating effect of emotion in this study by comparing the
opinion treatment effect across levels of embarrassment and pride separately and for all participants,
regardless of treatment. To illustrate, take the case of embarrassment. The meaning and effects of a high
level of embarrassment differ depending on whether it occurs in participants who received the pro-
individualism or the anti-individualism treatment. In response to the anti-individualism treatment, where
the participant is in the minority, embarrassment signals a participant is likely to conform, as expected;
however, in response to the pro-individualism treatment, the unusual circumstance of embarrassment in
response to being in the majority suggests a participant may instead deviate from the perceived norm
because he or she is, evidently, uncomfortable being in the mainstream. (A parallel, opposite, result would
occur if one concentrated on pride.) Thus, despite a focus on one emotion, the changing context means
such an analysis is not comparing like to like. Interacting the opinion treatment variable with the above-
described Self-Conscious Emotion Intensity scale better tests the emotion moderation hypothesis.
32 Randomization successfully balanced the treatment groups with respect to age and political ideology
but not sex or race. These two variables are therefore added to the analyses as controls; however, their
addition does not substantially alter the results of the analyses.

Ô Springer

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms
 240 Polit Behav (2015) 37:221-251

Table 3 ANOVA results for three models

Modell: Model 2: Model 3:

basic measured arousal
model emotion added

Opinion treatment .90 .72 .75

Emotion treatment .02

Opinion treatment x emotion treatment 4.46*

Self-conscious emotion intensity - 1.83 .36
Opinion treatment x self-conscious emotion intensity - 6.06* 4.51*
Emotional arousal - - 2.64

Opinion treatment x emotional arousal - - .01

Sex control .18 .22 .13

Race control 5.12* 5.65* 5.76*

N 82 82 81

Cells contain F-statistics and are marked

** p < .01; * p < .05; A p < .10

set-up allows for tests of HI (conformity) as well as H6 (emotion moderation).33 For

HI to be fully supported, participants' post-test views in the pro-individualism
treatment groups should be more "conservative" than participants' post-test views in
the anti-individualism groups. For H6 to be supported, this difference - evidence of
majority influence - should be greater when the relevant emotions are greater. In
other words, there should be a positive interaction between the opinion treatment and
self-conscious emotions. Finally, note that no main effect for emotion is expected.
I.e., pride or embarrassment in and of itself should not cause individualism to
increase or decrease.

The first ANOVA includes the opinion treatment variable (coded 0/1 ), the exogenous
emotion treatment variable (0/1), and their interaction. The results are displayed in the
first column of Table 3. With respect to the first factor, majority opinion did not appear
to influence participants' views on average F(l, 82) = .90, p = .35. However, the
interaction term is statistically significant: F(l, 82) = 4.46, p < .05, two-tailed.34 In
other words, the effect of the opinion stimuli changed when emotions were primed.
To clarify these relationships, experimental group means are graphed below in
Fig. 4. Without emotions primed, those in the anti-individualism treatment group
reported slightly more individualistic attitudes than those in the pro-individualism
treatment group, contrary to expectations; however, when pride or embarrassment
was primed, this relationship reversed. Participants appeared to conform to the
perceived majority, as expected.

33 Note that this study tests H6 in the context of conformity to in-group norms. Future research will be
needed to address the emotion moderation hypothesis with respect to polarization in response to out-
group norms.

34 While hypotheses with clear directional claims continue to be tested, the F ratio is akin to a two-tailed
test.

Ô Springer

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms
Polit Behav (2015) 37:221-251 241

Fig. 4 Predicted values for opinion and e

interaction between opinion and emotion t
Race to 0 (white)

One drawback of the above anal

whether or not participants re
measure of participants' subjec
evoked pride or embarrassment i
primes were not present. A seco
an emotion group with the Sel
Table 3 (second column). Here,
treatment, but the interaction
82) = 6.06, p < .05. The pattern of
See Fig. 5. No social influence ap
very low levels of self-conscious
advance up the emotion scale,
statistically significant at and ab

Study 2 Discussion

In this experiment, the self-con

to increase individuals' suscept
emerged for political conformit

35 Self-reported emotion is entered in

technically an ANCOVA (with interacti
actual range is 0-.75 because few particip

Ô Springer

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms
 242 Polit Behav (2015) 37:221-251

Fig. 5 Predicted values by opinion treatm

values for interaction between opinion t
controls. In figure, Sex is held to 1 (fe
estimates are statistically significant (p

information and among those re

tional reactions. However, when
those experiencing higher-than-
shift in the direction of majority
This said, one could suggest an i
self-conscious emotions increased
Perhaps emotions mattered, bu
example, maybe the emotional st
emotional arousal (Lazarus 1991,
opinion stimuli, and, therefore,
hypothesis by adding a measure o
the opinion treatment to the orig
emotion variables. See the third
supported by the analysis. The Op
F statistic of nearly 0. The F s
Intensity interaction declines slig
81) = 4.51, p < .05. Note that if e
Arousal measure is examined sep

36 An average of three basic emotions:

â Springer

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms
 Polit Behav (2015) 37:221-251 243

Discussion and Conclusion

Overall, the pattern of results observed reflects the expectations set out in the six
hypotheses. The findings from the first experiment demonstrate the power of in-group
norms to influence and out-group norms to polarize as well as the relevance of
gradations of subjective identity to these processes. The first experiment's findings also
suggest, but do not demonstrate, a causal role for self-conscious emotions. Results from
the second experiment directly support emotional causation. While these experimenta
effects will likely decay over time (e.g., see Chong and Druckman 2013), they also stem
from a single (impersonal) interaction with a group norm. Repeated interactions would
likely reinforce attitudes and lead to stabilization over time.
Alternative explanations for the pattern of findings are worth considering,
however. Could the pattern of results in the two experiments be due to the more
common argument that social influence, generally speaking, is due to information
exchange and/or argument? This is unlikely given that the stimuli deliberately did
not include any substantive information or arguments. What about other mor
"rational" explanations, such as the possibility that study participants were
conforming for self-interested reasons (Chong 2000) or because they constructed
arguments to explain the majority's view, persuading themselves in the proces
(Mutz 1998)? These theories are also unlikely to explain the experimental results.
Theories such as Chong' s that emphasize self-interested motivations tend to focus
on public behavior, not private opinions, as were the focus of both experimental
studies. With respect to Mutz' s theory, those mechanisms could explain the in-
group influence observed in the first experiment but not the empirical patterns
related to out-group norms or self-conscious emotion.
In sum, the empirical results support the "social-emotional influence" frame-
work proposed herewith. That framework provides an explanation for a phenom-
enon - political conformity - that has long been underexplored and not well
understood as a result. Likewise, it provides a window into between-group
polarization, a topic highly relevant to contemporary U.S. politics. Social identity
allows us to understand which groups will influence a person, and in what way.
Emotion helps us to understand why people fall into line with in-group expectations
so readily despite exposure to diverse perspectives on politics.
In addition, the social-emotional perspective on conformity and polarization
helps to make sense of certain puzzling aspects of public opinion. For example, it
offers an explanation as to why norms, values, and attitudes are affectively charged
(Banaji and Heiphetz 2010; Rokeach 1973) and closely correlated with group
boundaries. The framework also provides insight into why people tend to feel proud
of shared group norms and to feel shame when they disregard them. Finally, the
theoretical framework contributes to the growing literature on emotions and politics
Heretofore, most empirical emotions and politics scholars have focused on the basic
emotions of fear/anxiety, enthusiasm, and anger (e.g., Brader 2006; Gadarian 2010;
Huddy et al. 2007; Marcus et al. 2000; Valentino et al. 201 1), leaving self-conscious
and other more complex emotions to the side. Given the many ways in which group
identity, social relationships, and social status intersect with politics, further
examination of the political relevance of self-conscious emotions would be fruitful.

Ô Springer

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms
 244 Polit Behav (2015) 37:221-251

Acknowledgments A great many people

observations, criticisms, and suggestions
Ted Brader, Don Herzog, Don Kinder, an
Arceneaux, John Bullock, Eric Dickson,
Nathan Kalmoe, Chris Karpowitz, Skip
Preston, Lynn Sanders, Laura Stoker, and
Michigan for generous financial support
and community members in Michigan w

Appendix 1: Religious Identity

Stimuli

PLEASE READ THE TEXT BELOW CAREFULLY. WHEN YOU ARE DONE, ADVANCE TO THE
NEXT PAGE TO ANSWER SOME RELATED QUESTIONS

[Headline insert]
As you may know, the issue of "family values" continues to be discussed in the media. From time-to-
time, public opinions polls are carried out to find out what different types of Americans believe
regarding family values. [Body insert A] According to the survey:
[Body insert B]
What about you? We would like to know your opinion on family values.

Catholics Conservative Catholics Progressive Evangelicals Conservative

Condition Condition Condition

Headline Recent Polls Indicate Recent Polls Indicate Recent Polls Indicate
insert Catholics Are Strong Catholics Are Less Evangelicals Are Strong
Supporters of Family Supportive of Family Supporters of Family
Values Values Values

Body For example, one recent For example, one recent For example, one recent
insert A survey indicates that survey indicates that survey indicates that
American Catholics today American Catholics today American Evangelical (or
continue to strongly seem to question the "born again") Christians
support traditional family importance of traditional today are strong supporters
values family values of traditional family values
Body The majority of Catholics A majority of Catholics say A ma
insert B who marry stay married one can be a good Catholic say
and never divorce without obeying the avoided, even in the event
Church's teaching on of an unhappy marriage
divorce

A majority of Catholics A majority of Catholics say A majority of Evangelicals

oppose abortion one can be a good Catholic oppose abortion
without obeying the
Church's teaching on
abortion

A majority of Catholics A majority of Catholics say A majority of Evangelicals

oppose gay marriage the Church's opposition to oppose gay marriage
gay marriage is not very
important to them

^ Springer

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms
 Polit Behav (2015) 37:221-251 245

Measures37

Social Conservatism Scale

Divorce Subscale

Divorce in this country should be more difficult to obtain than it is now.

Divorce is usually the best solution when a couple can't seem to work out their
marriage problems.
Premarital Sex Subscale

It is wrong for a man and a woman to have sexual relations before marriage.
It's a good idea for a couple who intend to get married to live together first.

Gay Rights Subscale

Sexual relations between two adults of the same sex is wrong.

Homosexual couples should have the right to marry one another.

Sex and Young People Subscale

Sex education has no place in the nation's public schools.

Methods of birth control should be available to teenagers who need them.

Abortion Question

There has been discussion about abortion during recent years. Which one of the
opinions below best represents your view? By law, abortion should never be
permitted./The law should permit abortion only in the case of rape, incest, or
when the woman's life is in danger./The law should permit abortion for reasons
other than rape, incest, or danger to the woman's life, but only after the need for
the abortion has been clearly established./By law, a woman should always be able
to obtain an abortion as a matter of personal choice./Other

Catholic Identity Scale

How important is being Catholic to you?

How well does the term "Catholic" describe you?

Emotional Reactions to Stimuli

Did the information make you feel [proud/ashamed]?

37 Seven answer categories for Likert items ranged from "strongly agree" to "strongly disagree."
Identity items included five possible responses, from, e.g., "not very important" to "extremely
important." Emotion questions had five possible responses, from "not at all" to "extremely."

Ö Springer

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms
 246 Polit Behav (2015) 37:221-251

Appendix 2: Self-Conscious Em
(Study 2)

Pro-Individualism Stimulus

HOW DO YOU COMPARE?

In the spring of 2005, researchers at UCLA (University of Calif
Angeles) carried out opinion polls of college students at ten major u
throughout the United States, including The University of Michi
researchers asked random samples of undergraduate students at each un
about what majors they chose and why, about study habits and extr
activities, about Internet use, about their consumer habits, and, finally, a
social attitudes and political opinions.
Two of the survey questions focused on attitudes regarding "ec
individualism." According to results published last year in Public
Quarterly , most University of Michigan students agree with this princi
64 % of University of Michigan undergraduates either agreed or stron
with the statement "Individuals should strive to be financially self-reli

61 % of University of Michigan undergraduates either agreed or stron

with the statement "One ought to work hard in life."

19. You were asked to respond to the same two statements on page 2.
your answers to the published data and then check the appropriate res

Ô Springer

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms
 Polit Behav (2015) 37:221-251 247

I am in agreement with the ma

students....

Pride Stimulus

Take a moment to imagine each of the following scenarios, focusing on how each
situation would make you feel. Then circle the situation that you believe would
make you feel the best.

a. You leave school in April to spend the summer at home. One of your goals is to
improve the way you look - get in shape, buy some new clothes, maybe get a
new haircut, etc. When you return to school in the fall, everyone tells you how
great you look. You go to a party the first weekend back, and two cute guys (or
girls) approach you during the evening and ask you out.
b. You attend a family gathering over winter break with various family members.
One of your relatives asks you how school is going. As it happens, you got
straight As in the fall semester and have secured a really prestigious summer
internship, all of which you tell your relatives. The group gushes about your
accomplishments, and your mom looks especially pleased.
c. You are standing on the curb of a busy street, waiting for the light to turn green
so that you can cross, when you see a little girl wander away from her mother
and dart into the street. You run after her into the traffic, pick her up, and return
her to her mother. A small crowd that has gathered on the sidewalk to watch
breaks into applause.

Embarrassment Stimulus

Take a moment to imagine each of the following scenarios, focusing on how each
situation would make you feel. Then circle the situation that you believe would
make you feel the worst.

a. You are on a first date with someone you really like. You go to dinner, then to a
party. As the evening is coming to an end, both of you are sitting together on a
couch. Your date leans in close to you, and you're thinking it is finally time for
a kiss. But, instead, your date whispers to you, "Sorry to tell you this, but, uh,
the zipper on your pants has been down since we left the restaurant."
b. It's a warm spring day, and you are walking through the Diag, which is filled
with students socializing, studying, playing Frisbee, etc. All of a sudden you
trip and, with a loud grunt, fall down. Several books and the bag you had been

£) Springer

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms
 248 Polit Behav (2015) 37:221-251

carrying scatter all around you.

are doing to stare at you sprawle
c. You are attending the weddin
quiet, except for the bride and g
you get a case of the hiccups.
around you turn to you and say "shhhhh

mouth, but you can't stop hiccupping.

Dependent Measures

The federal government currently gives money fo

high school graduates. Some people believe that the
to those low-income graduates who have taken r
Suppose these people are at one end of a scale, at p
such college grants should go to all low-income hig
what courses they have taken. Suppose these people
And, of course, some other people have opinions so
2-6. Where would you place yourself on this scale?
Currently, the government in Washington provid
mothers who have dependent children; this prog
"welfare." Some people feel that the government s
work in order to receive welfare benefits. Suppose t
scale, at point 1 . Others feel that the government
of work status. Suppose these people are at the othe
some other people have opinions somewhere in
would you place yourself on this scale?

References

Abrams, D., Wethereil, M., Cochrane, S., Hogg, M. A., & Turner, J. C. (1990). Knowing what to think by
knowing who you are: Self-categorization and the nature of norm formation, conformity and group
polarization. British Journal of Social Psychology, 29, 97-119.
Alwin, D. F., Cohen, R. L., & Newcomb, T. M. (1991). Political attitudes over the lifespan: The
Bennington women after fifty years. Madison, WI: University of Wisconsin Press.
Asch, S. E. (1951). Effects of group pressure upon the modification and distortion of judgments. In H.
Guetzkow (Ed.), Groups, leadership, and men: Research in human relations. Pittsburgh, PA:
Carnegie Press.
Banaji, M. R., & Heiphetz, L. (2010). Attitudes. In S. T. Fiske, D. T. Gilbert, & G. L. Lindzey (Eds.),
Handbook of social psychology (5th ed.). New York, NY: Wiley.
Beck, P. A. (2002). Encouraging political defection: The role of personal discussion networks in partisan
desertions to the opposition party and Perot votes in 1992. Political Behavior, 24(4), 309-337.
Berelson, B. R., Lazarsfeld, P. F., & McPhee, W. N. (1954). Voting: A study of opinion formation in a
presidential campaign. Chicago, IL: University of Chicago Press.
Blanchette, I., & Richards, A. (2010). The influence of affect on higher level cognition: A review of
research on interpretation, judgment, decision making and reasoning. Cognition and Emotion, 24(4),
561-595.

Ô Springer

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms
 Polit Behav (2015) 37:221-251 249

Bolsen, T. (2013). A light bulb goes on:

Behavior, 35, 1-20.
Brader, T. (2006). Campaigning for heart
Chicago, IL: University of Chicago Pres
Brown, R. (2000). Social identity theory:
European Journal of Social Psychology
Bullock, J. G., Green, D. P., & Hal, S. E. (
answer). Journal of Personality and So
Campbell, A., Converse, P. E., Miller, W.
ed.). Chicago, IL: University of Chicago
Chong, D. (2000). Rational lives: Norms a
Chicago Press.
Chong, D., & Druckman, J. N. (2013). Cou
Clark, R. D., & Maass, A. (1988). The role
minority influence. European Journal o
D'Antonio, W. V. (2007). American Cath
Lanham, MD: Rowman & Littlefield.
Davis, J. A., Smith, T. W., & Marsden, P. V. (2008). General social surveys, 1972-2008: Cumulative
codebook. Chicago, IL: National Opinion Research Center.
Deutsch, M., & Gerard, H. B. (1965). A study of normative and informational social influences upon
individual judgment. In H. Proshansky & B. Seidenberg (Eds.), Basic studies in social psychology.
New York, NY: Holt, Rinehart and Winston.
Djupe, P. A., & Gilbert, C. P. (2008). The political influence of churches. Cambridge: Cambridge
University Press.
Downs, A. (1957). An economic theory of democracy. New York, NY: Harper and Row.
Elster, J. (1999). Alchemies of the mind: Rationality and the emotions. Cambridge: Cambridge University
Press.

Feldman, S. (1999). Economic values and inequality. In J. P. Robinson, P. R. Shaver, & L. S. Wrightsman
(Eds.), Measures of political attitudes. San Diego, CA: Academic Press.
Feldman, S., & Johnston, C. (2014). Understanding the determinants of political ideology: Implications of
structural complexity. Political Psychology, 35(3). doi: 10.1 111/pops.l 2055.
Festinger, L. (1950). Informal social communication. Psychological Review, 57, 271-282.
Gadarian, S. K. (2010). The politics of threat: How terrorism news shapes foreign policy attitudes. The
Journal of Politics, 72(2), 469-483.
Gerber, A. S., Green, D. P., & Larimer, C. W. (2008). Social pressure and vote turnout: Evidence from a
large-scale field experiment. American Political Science Review, 702(1), 19-31.
Gerber, A. S., Green, D. P., & Larimer, C. W. (2010). An experiment testing the relative effectiveness of
encouraging voter participation by inducing feelings of pride or shame. Political Behavior, 32,
409-422.
Gerber, A. S., & Rogers, T. (2009). Descriptive social norms and motivation to vote: Everybody's voting
and so should you. The Journal of Politics, 77(1), 178-191.
Haslam, S. A., Ellemers, N., Reicher, S. D., Reynolds, K. J., & Schmitt, M. T. (2010). The social identity
perspective today: An overview of its defining ideas. In T. Postmes & N. R. Branscombe (Eds.),
Rediscovering social identity : Key readings. New York, NY: Psychology Press.
Hogg, M. A., & Reid, S. A. (2006). Social identity, self-categorization, and the communication of group
norms. Communication Theory, 16, 7-30.
Huckfeldt, R., & Sprague, J. (1987). Networks in context: The social flow of political information.
American Political Science Review, 81, 1197-1216.
Huckfeldt, R., & Sprague, J. (1995). Citizens, politics, and social communication: Information and
influence in an election campaign. Cambridge: Cambridge University Press.
Huddy, L. (2001). From social to political identity: A critical examination of social identity theory.
Political Psychology, 22(1), 127-156.
Huddy, L., Feldman, S., & Cassese, E. (2007). On the distinct political effects of anxiety and anger. In W.
R. Neuman, G. E. Marcus, M. MacKuen, & A. N. Crigler (Eds.), The affect effect: Dynamics of
emotion in political thinking and behavior. Chicago, IL: University of Chicago Press.
Huddy, L., & Khatib, N. (2007). American patriotism, national identity, and political involvement.
American Journal of Political Science, 51, 63-77.

Ô Springer

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms
 250 Polit Behav (2015) 37:221-251

Imai, K., Tingley, D., & Yamamoto, T. (2

Journal of the Royal Statistical Society
Lazarus, R. S. (1991). Emotion and adapt
Lazer, D. (2011). Networks in political
44( 1), 61-68.
Leege, D. C., Wald, K. D., Krueger, B. S.,
Social change and voter mobilization
Princeton University Press.
Lewis, M. (2000). Self-conscious emotion
M. Haviland-Jones (Eds.), Handbook of
Luhtanen, R., & Crocker, J. (1992). A collective self-esteem scale: Self-evaluation of one's social
identity. Personality and Social Psychology Bulletin , 18, 302-318.
Mackie, D. M., Maitner, A. T., & Smith, E. R. (2009). Intergroup emotions theory. In T. D. Nelson (Ed.),
Handbook of prejudice, stereotyping, and discrimination. New York, NY: Psychology Press.
Marcus, G. E., Neuman, W. R., & MacKuen, M. (2000). Affective intelligence and political judgment.
Chicago, IL: University of Chicago Press.
Marques, J. M., Abrams, D., & Serôdio, R. S. (2001). Being better by being right: Subjective group
dynamics and derogation of in-group deviants when generic norms are undermined. Journal of
Personality and Social Psychology, 81(3), 436-447.
McClurg, S. D. (2006). The electoral relevance of political talk: Examining disagreement and expertise
effects in social networks on political participation. American Journal of Political Science, 50(3),
737-754.

McNair, M. L., & Droppleman, L. F. (1971). Profile of mood states. San Diego, CA: Educational an
Industrial Testing Service.
Milgram, S. (1992). The individual in a social world: Essays and experiments (2nd ed.). New York, NY
McGraw-Hill Inc.
Miller, R. S. (2007). Is embarrassment a blessing or a curse? In J. L. Tracy, R. W. Robins, & J.
P. Tangney (Eds.), The self-conscious emotions: Theory and research. New York, NY: Guilford
Press.

Morton, R. B., & Williams, K. C. (2010). Experimental political science and the study of causality: From
nature to the lab. Cambridge: Cambridge University Press.
Moskalenko, S., McCauley, C., & Rozin, P. (2006). Group identification under conditions of threat:
College students' attachment to country, family, ethnicity, religion, and university before and after
September 11, 2001. Political Psychology, 27(1), 77-97.
Mutz, D. C. (1998). Impersonal influence: How perceptions of mass collectives affect political attitudes.
Cambridge: Cambridge University Press.
Nadeau, R., Cloutier, E., & Guay, J.-H. (1993). New evidence about the existence of a bandwagon effect
in the opinion formation process. International Political Science Review, 14(2), 203-213.
Nelson, T. E., Gwiasda, G., & Lyons, J. (2011). Vilification and values. Political Psychology, 32(5),
813-835.
Nickerson, D. W. (2008). Is voting contagious? Evidence from two field experiments. The American
Political Science Review, 102(1), 49-57.
Noelle-Neuman, E. (1993). The spiral of silence: Public opinion - our social skin (2nd ed.). Chicago, IL:
University of Chicago Press.
Panagopoulos, C. (2010). Affect, social pressure and prosocial motivation: Field experimental evidence
of the mobilizing effects of pride, shame and publicizing behavior. Political Behavior, 32, 369-386.
Panksepp, J. (1994). The basics of basic emotions. In P. Ekman & R. J. Davidson (Eds.), The nature of
emotion: Fundamental questions. New York, NY: Oxford University Press.
Pew Research Center. (2008). A portrait of American Catholics on the eve of Pope Benedict's visit.
Washington, DC: Pew Research Center Publications.
Putnam, R., & Campbell, D. E. (2010). American grace: How religion divides and unites us. New York,
NY: Simon & Schuster.
Rokeach, M. (1973). The nature of human values. New York, NY: The Free Press.
Scheff, T. (1988). Shame and conformity: The deference-emotion system. American Sociological Review
53, 395-406.
Sears, D. O., & Levy, S. (2003). Childhood and adult political development. In D. O. Sears, L. Huddy, &
R. Jervis (Eds.), Oxford handbook of political psychology. Oxford: Oxford University Press.
Sherif, M. ([1936] 1966). The psychology of social norms. New York, NY: Harper & Row.

Springer

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms
 Polit Behav (2015) 37:221-251 251

Sinclair, B. (2012). The social citizen: Peer

Chicago Press.
Small, D. A., & Lerner, J. S. (2005). Emotio
a welfare case. Political Psychology, 29,
Sokhey, A. E., & Djupe, P. A. (2011). Int
Science & Politics, 44( 1), 55-59.
Tajfel, H. (1978). Social categorization,
Differentiation between groups : Studies
Academic Press.

Tajfel, H., & Turner, J. C. (1986). The social identity theory of intergroup behavior. In W. G. Austin & S.
Worchel (Eds.), Psychology of intergroup relations. Chicago, IL: Nelson-Hall Publishers.
Terry, D. J., & Hogg, M. A. (1996). Group norms and the attitude-behavior relationship: A role for group
identification. Personality and Social Psychology Bulletin, 22(8), 776-793.
Terry, D. J., Hogg, M. A., & White, K. M. (1999). The theory of planned behaviour: Self-identity, social
identity, and group norms. British Journal of Social Psychology, 38( 3), 225-244.
Theiss-Morse, E. (2009). Who counts as an American? The boundaries of national identity. Cambridge:
Cambridge University Press.
Turner, J. C. (1985). Social categorization and the self-concept: A social cognitive theory of group
behavior. In E. Lawler (Ed.), Advances in group processes: Theory and research. Greenwich, CT:
JAI Press.

Turner, J. C. (1991). Social influence. Bristol, PA: Open University Press.

Turner, J. C., with Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. S. (1987). Rediscovering
the social group: A self -categorization theory. Oxford: Blackwell.
Valentino, N. A., Brader, T., Groenendyk, E. W., Gregorowicz, K., & Hutchings, V. (2011). Election
night's alright for fighting: The role of emotions in political participation. The Journal of Politics,
73(1), 156-170.
Wald, K. D., Owen, D. E., & Hill, S. S., Jr. (1988). Churches as political communities. The American
Political Science Review, 82, 531-548.
Walsh, K. C. (2003). Talking about politics: Informal groups and social identity in American life.
Chicago, IL: University of Chicago Press.
Wood, W., Pool, G. J., Leck, K., & Purvis, D. (1996). Self-definition, defensive processing, and influence:
The normative impact of majority and minority groups. Journal of Personality and Social
Psychology, 71, 1181-1193.
Zajonc, R. B. (1980). Feeling and thinking: Preferences need no inferences. American Psychologist, 35,
151-175.

Springer

This content downloaded from

104.207.138.102 on Mon, 29 Apr 2024 16:30:12 +00:00
All use subject to https://about.jstor.org/terms